Implementación de un detector de coral utilizando filtros Gabor Wavelets y máquinas de aprendizaje
Abstract
RESUMEN
Este trabajo se enfoca en la implementación de un detector de arrecife de coral de desempeño rápido que se utiliza para un vehículo autónomo submarino (Autonomous Underwater Vehicle, AUV, por sus siglas en inglés). Una detección rápida de la presencia de coral asegura la estabilización del AUV frente al arrecife en el menor tiempo posible, evitando colisiones con el coral. La detección de coral se hace en una imagen que captura la escena que percibe la cámara del AUV. Se realiza una clasificación píxel por píxel entre dos clases: arrecife de coral y el plano de fondo que no es coral. A cada píxel de la imagen se le asigna un vector característico, el mismo que se genera mediante el uso de filtros Gabor Wavelets. Éstos son implementados en C++ y la librería OpenCV. Los vectores característicos son clasificados a través de nueve algoritmos de máquinas de aprendizaje. El desempeño de cada algoritmo se compara mediante la precisión y el tiempo de ejecución. El algoritmo de Árboles de Decisión resultó ser el más rápido y preciso de entre todos los algoritmos. Se creó una base de datos de 621 imágenes de corales de Belice (110 imágenes de entrenamiento y 511 imágenes de prueba).
Palabras clave: AUV, arrecife de coral, máquinas de aprendizaje, filtros Gabor Wavelets, OpenCV.
ABSTRACT
This work focuses on the implementation of a fast coral reef detector that is used for an Autonomous Underwater Vehicle (AUV, its acronym in English). A fast detection of the presence of coral ensures the AUV stabilization in front of coral reef in the shortest possible time, avoiding collisions with coral. The coral detection is carried out on an image that captures the scene that the AUV’s camera perceives. A pixel-by-pixel classification is performed between two classes: coral reef and the background that is non-coral reef. Each pixel of the image is assigned to a feature vector, which is generated by using Gabor Wavelet filters. These are implemented in C++ and the OpenCV library. The feature vectors are classified using nine machine learning algorithms. The performance of each algorithm is compared with the accuracy and execution time. The Decision Tree algorithm proved to be the fastest and most accurate of all the algorithms. We created a database of 621 images of coral reefs in Belize (110 of training images and 511 of testing images).
Keywords: AUV, coral reef, machine learning, Gabor Wavelets filters, OpenCV.
Downloads
Metrics
References
Bilmes, J., 1998. A gentle tutorial of the EM algorithm and its application to parameter estimation for Gaussian mixture and hidden Markov models. International Computer Science Institute and Computer Science Division, 10 pp. Disponible en http://crow.ee.washington.edu/people/ bulyko/papers/em.pdf.
Bradski, G., A. Kaehler, 2008. Learning OpenCV: Computer Vision with the OpenCV Library, 579 pp. O’Reilly Media Publ., eBook, disponible en: http://it-ebooks.info/book/299/.
Chen, L., G. Lu, D. Zhang, 2004. Effects of different gabor filter parameters on image retrieval by texture. Proceedings 10th International Conference on Multimedia Modelling, IEEE Publ., 273-278.
Coralbot, T., 2013. About coral reefs and current approaches for restoration. Obtenido de Coralbot ; http://www.coralbots.org/.
Friedman, J., 2000. Greedy function approximation: A gradient boosting machine. Annals of Statistics, 1189-1232. Disponible en http://statweb.stanford.edu/~jhf/ftp/trebst.pdf.
Geurts, P., D. Ernst, L. Wehenkel, 2006. Extremely randomized trees. Springer Science + Business Media, 40 pp. Disponible en http://www.montefiore.ulg.ac.be/~ernst/uploads/news/id63/ extremely-randomized-trees.pdf.
Johnson-Roberson, M., S. Kumar, S. Willams, 2006. Segmentation and classification of coral for oceanographic surveys: A semi-supervised machine learning approach. Proceedings of OCEANS-Asia Pacific, 16-19 May, Singapore, 6 pp.
Loh, W.-Y., 2011. Classication and regression tree. WIREs Data Mining and Knowledge Discovery, John Willey & Sons Inc., 14-23. Disponible en http://www.stat.wisc.edu/~loh/treeprogs/ guide/wires11.pdf.
Luber, M., L. Spinello, K.O. Arras, 2011. People tracking in rgb-d data with on-line boosted target models. In: Proceedings of the IEEE/RSJ Int. Conf. on Intelligent Robots and Systems, 6 pp. Disponible en http://www2.informatik.uni-freiburg.de/~spinello/luberIROS11.pdf.
Manjunath, B.S., W.Y. Ma, 1996. Texture features for browsing and retrieval of image data. IEEE Transactions on Pattern Analysis and Machine Intelligence, 18(8), 837-842.
Marcos, M.S., L. David, E. Peñaflor, V. Ticzon, M. Soriano, 2008. Automated benthic counting of living and non-living components in Ngedarrak Reef, Palau via subsurface underwater video. Environmental Monitoring and Assessment, 145, 177-184.
Mitchell, T., 1997. Machine learning. McGraw-Hill, Inc., New York, USA, 414 pp.
Ojala, T., M. Pietikäinen, T. Mäenpää, 2002. Multiresolution gray-scale and rotation invariant texture classification with local binary patterns. IEEE Transactions on Pattern Analysis and Machine Intelligence, 24(7), 971-987.
Ontrup, J., H. Ritter, H. Wersing, 2004. A computational feature binding model of human texture perception. Cognitive Processing, 34-44.
Purser, A., M. Bergmann, T. Lundälv, J. Ontrup, T.W. Nattkemper, 2009. Use of machine-learning algorithms for the automated detection of cold-water coral habitats: a pilot study. Inter-Research, Marine Ecology Progress Series, Vol. 397, 241-251.
Stough, J., L. Greer, B. Matt, 2012. Texture and color distribution-based classification for live coral detection. Proceedings of the 12th International Coral Reef Symposium, 9-13 July, Cairns, Australia, 5 pp. Disponible en http://www.icrs2012.com/proceedings/manuscripts/ ICRS2012_5D_2.pdf.
Downloads
Published
How to Cite
Issue
Section
License
Copyright © Autors. Creative Commons Attribution 4.0 License. for any article submitted from 6 June 2017 onwards. For manuscripts submitted before, the CC BY 3.0 License was used.
You are free to:
 |
Share — copy and redistribute the material in any medium or format |
 |
Adapt — remix, transform, and build upon the material for any purpose, even commercially. |
Under the following conditions:
 |
Attribution — You must give appropriate credit, provide a link to the licence, and indicate if changes were made. You may do so in any reasonable manner, but not in any way that suggests the licenser endorses you or your use. |
| No additional restrictions — You may not apply legal terms or technological measures that legally restrict others from doing anything the licence permits. |
